Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.mel.connect.com.au!munnari.OZ.AU!spool.mu.edu!howland.reston.ans.net!nntp.coast.net!news.dacom.co.kr!news.kreonet.re.kr!overload.lbl.gov!eubie.lbl.gov!jin From: jin@eubie.lbl.gov (Jin Guojun[ITG]) Newsgroups: comp.unix.bsd.freebsd.misc Subject: Re: Parity SIMMS really necessary? Date: 7 Dec 1995 20:55:07 GMT Organization: Lawrence Berkeley Laboratory Lines: 55 Message-ID: <4a7kbb$8uj@overload.lbl.gov> References: <49lbnr$4fq@interport.net> <49qabp$efi@zuul.nmti.com> <4a2fug$b0u@cnn.nas.nasa.gov> NNTP-Posting-Host: eubie.lbl.gov In article <4a2fug$b0u@cnn.nas.nasa.gov>, Keith C. Thompson <thompson@sun446.nas.nasa.gov> wrote: >In article <49qabp$efi@zuul.nmti.com> peter@nmti.com (Peter da Silva) writes: >>In article <49lbnr$4fq@interport.net> David Tay <davidtay@interport.net> wrote: >>> I'm currently using FreeBSD with parity simms. There's a $50-$100 >>> difference with parity SIMMs. I would like to save some money and use >>> non-parity SIMMs. >> >>I'm obviously naive. I had no idea they even *made* non-parity SIMMs. >> >>The whole idea of putting that much memory in a box without at *least* >>parity (and preferably ECC) makes me sick in the stomach. > >It's gotten worse. There are few true parity SIMMs being sold these days. >Almost all of the "parity" SIMMs that I have seen in the big computer >places around the Bay Area are non parity SIMMs with a parity generator. >It really sucks when your mother board uses the opposite parity..... > >At least the BIOS on most mother boards these days can be configured >to ignore the parity. This allows one to use non parity SIMMs. > >- Keith This is correct. The thing is going up and down. The new motherboards are made to run at NO-parity environment. The silicon technology is better and better. Current memory chips are very reliable. Once they are installed, they almost live forever, unless you overheat or throw electrical static on them. Back to 15-20 years ago, parity is not required for memory. So for the later. I heard many people had memory problems. However, I do not know how many of them really are memory chip defecting. In the other word, once the memory failed, it will not be able to be used again. Are they? Many problem are caused by vendor who made memory to do the marginal job. For example, 66Mhz and 100 MHz CPU request 60ns memory. In my experience, Almost more than halt of the vendors use 70ns chip instead. Most memory chip design may have 5%-8% marginal ability and won't work in this way. Only a few memory makers make memory chip able to work in high range (clock). Therefore, if the correct memory chip is used, the parity is not necessary because the BIOS will check the memory at boot time to see if any defects. Then during the runtime, the new memory tecnology will provide a safe live if no overheating and electrical static applied. Remember, the most memory checking program will not detect marginal working. The MS-DOS2.2 HIMEMORY test will do 95% jobs. That is, if your memory can pass MS-DOS2.2 HiMEM test, it almost guaranties your memory is OK. However, do not try to let memory do the marginal job is the key. Use 60ns chip for 16.7 MHz or lower bus clock, and 70ns chip for 15 MHz or lower bus clock. Good Luck, -- /-------------- Jin Guojun ------------ v ---- Internet: g_jin@lbl.gov ----\ | Imaging & Distributed Computing | Usenet: ucbvax!g_jin@lbl.gov | | Lawrence Berkeley Laboratory | Bitnet: -- | | 50B-2239, Berkeley, CA 94720 - jin%george.lbl.gov@Csa3.LBL.Gov |